--- title: Langfuse Assistant description: Explore your Langfuse project data and take actions in Langfuse from inside the UI. sidebarTitle: Assistant --- # Langfuse Assistant **Where is this feature available?** | Plan | Availability | | --- | --- | | Hobby | Available | | Core | Available | | Pro | Available | | Enterprise | Available | | Self Hosted | Not Available | The Langfuse Assistant is an in-product assistant for exploring your Langfuse project data and taking actions in Langfuse with your approval. You can ask about traces, observations, sessions, and metrics, or ask it to create datasets, dashboards, and other resources. Every run happens in the background. You get a notification when the Assistant is done or needs your input or approval. ## What you can do [#what-you-can-do] Ask the Assistant about your data and about Langfuse itself: - "Which traces had the highest latency yesterday?" - "Show me failed generations in the last hour." - "What's my token spend this week, broken down by model?" - "How do I set up an evaluator?" Or ask it to do the work: - "Create a dataset from the failed generations of last week." - "Build a dashboard of error rate and p95 latency for the last 7 days." - "Review this prompt against current best practices and propose a tighter version." - "Fetch a few hundred failed observations from the last 7 days and cluster the main errors with code." - "Set up a numeric score config for friendliness." - "Open the traces table filtered to errors from today." ## How it works [#how-it-works] The Assistant uses the Langfuse MCP server and additional tools to answer questions, inspect project data, and help you take actions in Langfuse. The Assistant can: - Query traces, observations, sessions, and metrics through Langfuse tools - Fetch large observation sets and process them with code in a sandbox, instead of loading every payload into the model context - Search Langfuse documentation - Propose links to Langfuse pages such as traces, sessions, dashboards, prompts, datasets, experiments, evals, alerts, and project settings - Create or update Langfuse resources such as datasets, dashboards, widgets, prompts, and score configs The Assistant is designed to be safe by default: - If a request would change data or configuration, the Assistant pauses and asks you to approve that specific action first. - If you choose **Always approve** for a tool, the Assistant reuses that tool without asking again. The choice applies to that conversation only. The Assistant can only use actions that your account already has access to in the current project. Approval adds a second safety check, but it does not grant any extra permissions. The Assistant is also aware of your current context in the Langfuse UI, such as the current trace, observation, session, or dashboard, as well as user metadata such as your name, browser language, and timezone. This context is used to provide more relevant answers and propose actions. ## Background runs [#background-runs] Every Assistant run happens in the background, independently of your browser. Keep using Langfuse, or start another conversation, while a run is still working. You get a notification when it finishes, fails, or needs your input or approval. For repeating work in one conversation, choose **Always approve** on a tool you are comfortable leaving unattended. The Assistant then uses that tool for the rest of that conversation without asking again. ## Data privacy and security [#data-privacy-and-security] The Assistant follows the same data handling model as other Langfuse AI features. See [AI Features](/security/ai-features) for the canonical security and privacy details. Key points: - Model requests are handled in the same AWS region as your [Langfuse data region](/security/data-regions). - Assistant access is scoped to the authenticated user and current project. - Actions that can change data require your approval and remain limited to the permissions of the signed-in user. - AI feature tracing for product and service improvement can be enabled or disabled separately by organization admins and owners. ## Conversation history [#conversation-history] Assistant conversations are saved per user and project so you can return to previous conversations. ## Feedback [#feedback] You can submit thumbs-up or thumbs-down feedback on Assistant responses. If AI feature tracing is enabled, feedback may be used for product and service improvement. You can also share your thoughts in our [feedback discussion](https://github.com/orgs/langfuse/discussions/14196). ## Limitations [#limitations] The Assistant may misunderstand a question or return incomplete results. Verify important answers and review action details carefully before acting on them. ## Related resources [#related-resources] - [AI Feature Security](/security/ai-features) - [Langfuse MCP Server](/docs/api-and-data-platform/features/mcp-server) - [Langfuse Agent Skill](/docs/api-and-data-platform/features/agent-skill) - [Langfuse Assistant announcement](/changelog/2026-06-19-langfuse-assistant-public-beta) --- ## Agent Instructions This page is part of the [Langfuse](https://langfuse.com) documentation, published as plain Markdown for AI agents. Every page is available as Markdown by appending `.md` to its URL, or by sending an `Accept: text/markdown` header. This page: `https://langfuse.com/docs/langfuse-assistant.md`. ### Querying these docs If the answer is not on this page, query the documentation instead of guessing: - **Semantic search** across all Langfuse docs, returning an answer with the relevant pages and excerpts. Ask a specific, self-contained question: ```bash curl -sG "https://langfuse.com/api/search-docs" --data-urlencode "query=How do I trace a LangGraph agent?" ``` - **Index of every page**: , with per-section indexes [llms-docs.txt](https://langfuse.com/llms-docs.txt), [llms-integrations.txt](https://langfuse.com/llms-integrations.txt), and [llms-self-hosting.txt](https://langfuse.com/llms-self-hosting.txt). ### Before writing Langfuse code - **Install the [Langfuse Agent Skill](https://langfuse.com/docs/api-and-data-platform/features/agent-skill).** It encodes Langfuse's own best practices for instrumentation, prompt management, and evaluation, and materially improves results. - **Read [What does a good trace look like?](https://langfuse.com/docs/observability/best-practices.md)** before instrumenting an application. - **Verify endpoints, parameters, and response fields** against the [API reference](https://api.reference.langfuse.com) instead of inferring them from code examples. - **Use the [Langfuse CLI](https://langfuse.com/docs/api-and-data-platform/features/cli)** (`npx langfuse-cli api `) to read or write traces, prompts, datasets, and scores from the terminal. Found an error in these docs? Please open an issue at .